Using ring records and field surveys to predict the winter distribution of a migratory passerine

Abstract

Capsule: Ring records can be used to predict spatial changes in habitat suitability in wintering grounds of migratory birds as shown by a case study with the Meadow Pipit (Anthus pratensis) Aims:To forecast the winter distribution of the Meadow Pipit in the Iberian Peninsula and the Maghreb (1 700 000 km2 ) by combining ring records and fields surveys. Methods: We processed winter ring records with Maxent, a popular machine-learning technique used to predict the occurrence probability of species. We then tested whether the occurrence probabilities predicted by the model were significantly positively related to the actual abundance of the Pipit in Spain and Morocco. Results: Occurrence probabilities provided by Maxent explained 6% of Pipit abundance distribution in partial least squares regression analyses. There was, in addition, a weak effect of geographical location (1.5%) and a strong effect of local vegetation (17.6%) on Pipit distribution. The occurrence probabilities were used to classify the winter range of the Meadow Pipit in sectors of high, medium and low suitability. The most suitable sector occupied 9% of the region (153 000 km2 ) and was restricted to warm and moist lowlands where Pipits gather in grasslands. Conclusion: These results support the potential usefulness of cartographic models derived from ring records to describe the wintering range of migratory birds.